A Coregulatory Network of NR2F1 and Microrna-140

A Coregulatory Network of NR2F1 and microRNA-140

David Y. Chiang1,2☯, David W. Cuthbertson3☯, Fernanda R. Ruiz4, Na Li1*, Fred A. Pereira3,4*

1 Department of Molecular Physiology and Biophysics, Baylor College of Medicine, Houston, Texas, United States of America, 2 Interdepartmental Program in Translational Biology and Molecular Medicine, Baylor College of Medicine, Houston, Texas, United States of America, 3 Bobby R. Alford Department of Otolaryngology- Head and Neck Surgery, Baylor College of Medicine, Houston, Texas, United States of America, 4 Huffington Center on Aging and Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, Texas, United States of America

Abstract

Background: Both nuclear receptor subfamily 2 group F member 1 (NR2F1) and microRNAs (miRNAs) have been shown to play critical roles in the developing and functional inner ear. Based on previous studies suggesting interplay between NR2F1 and miRNAs, we investigated the coregulation between NR2F1 and miRNAs to better understand the regulatory mechanisms of inner ear development and functional maturation. Results: Using a bioinformatic approach, we identified 11 potential miRNAs that might coregulate target genes with NR2F1 and analyzed their targets and potential roles in physiology and disease. We selected 6 miRNAs to analyze using quantitative real-time (qRT) -PCR and found that miR-140 is significantly down-regulated by 4.5-fold (P=0.004) in the inner ear of NR2F1 knockout (Nr2f1–/–) mice compared to wild-type littermates but is unchanged in the brain. Based on this, we performed chromatin-immunoprecipitation followed by qRT-PCR and confirmed that NR2F1 directly binds and regulates both miR-140 and Klf9 in vivo. Furthermore, we performed luciferase reporter assay and showed that miR-140 mimic directly regulates KLF9-3’UTR, thereby establishing and validating an example coregulatory network involving NR2F1, miR-140, and Klf9. Conclusions: We have described and experimentally validated a novel tissue-dependent coregulatory network for NR2F1, miR-140, and Klf9 in the inner ear and we propose the existence of many such coregulatory networks important for both inner ear development and function.

Citation: Chiang DY, Cuthbertson DW, Ruiz FR, Li N, Pereira FA (2013) A Coregulatory Network of NR2F1 and microRNA-140. PLoS ONE 8(12): e83358. doi:10.1371/journal.pone.0083358 Editor: Bruce Riley, Texas A&M University, United States of America Received October 3, 2013; Accepted November 11, 2013; Published December 3, 2013 Copyright: © 2013 Chiang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was funded by grants T32AG00183 (to FRR) from the National Institutes of Health, 12BGIA12050207 (to NL) and 12PRE11700012 (to DYC) from the American Heart Association, and DC04585 and support from the Huffington Center on Aging and Department of Otolaryngology-HNS (to FAP). DYC was also supported by the Baylor College of Medicine Medical Scientist Training Program Caskey Scholarship. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. * E-mail: [email protected] (FAP); [email protected] (NL) ☯ These authors contributed equally to this work.

Introduction guidance, neocortical patterning, morphogenesis, and patterning of the cochlear sensory epithelium and the inner and Nuclear receptors are a large family of cellular regulators that outer hair cells in the organ of Corti [7–10]. A recent forward function as activators, repressors and silencers of transcription genetics screen in patients with congenital anomalies and when bound by specific ligands or cofactors [1,2]. They are cytogenetically balanced chromosomal rearrangements structured with a ligand-binding domain in the C-terminus, identified a critical paracentric microdeletion of the Nr2f1 locus, activation function domains, and a DNA-binding domain in the implicating haploinsufficiency of NR2F1 as a cause for a 4- N-terminus which binds to specific hormone response DNA year-old child’s deafness, dysmorphism and developmental elements in the genome [3]. Two major subclasses of nuclear delay [11]. receptors include the nuclear hormone receptors that respond In an effort to define transcription factor binding sites and to hormone ligands and the orphan nuclear receptors for which target genes in vivo, Montemayor et al. developed a endogenous ligands are not known [4]. Nuclear receptor methodology to identify NR2F1 target genes by intersecting subfamily 2 group F member 1 (NR2F1) is an orphan nuclear gene expression profiling, computational binding site queries, receptor required for the development of the inner ear and and evolutionary conservation data [12]. This approach is in cerebral cortex [5–8]. This is illustrated by loss-of-function in part based on the concept of ‘phylogenetic footprinting,’ which knockout mice (Nr2f1–/–) which impacts neurogenesis, axonal assumes evolution selects against mutations within DNA

PLOS ONE | www.plosone.org 1 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

regions that have an important regulatory function; thus regenerating or rejuvenating hair cells and neurons in patients creating evolutionary cold spots within the critical genomic with partial or complete hearing loss [25,26]. This study segments that regulate gene expression [13]. Using this discovered coregulatory interactions between NR2F1 and approach we identified and validated several direct target miRNA-140 in their regulation of the gene Krüppel-like factor 9 genes of NR2F1, and began to uncover regulatory and (Klf9). This is the first description of such a pathway and feedback networks at the transcriptional and post-translational delineates the regulatory actions of a circuit involving the non- level that might be necessary in the development and steroidal nuclear receptor transcription factor and miRNAs. physiology of mammals. To date, relatively few coregulatory networks have been delineated for interactions between Materials and Methods nuclear receptors and microRNAs (miRNAs), and even fewer have been experimentally validated [14]. Ethics Statement MicroRNAs were discovered as small temporal RNAs that The Institutional Animal Care and Use Committee of Baylor regulate developmental transitions in C. elegans. They have College of Medicine approved all studies involving the use of diverse expressive patterns and regulate a variety of areas, animals as mandated by the United States federal government. including embryologic development, physiology, pathophysiology, and growth, development, progression, and Identification of the 11 miRNAs and their putative metastasis in cancer [15–17]. Currently, there are 2,578 targets (human) and 1908 (mouse) mature unique miRNAs identified In order to find the overlap of any miRNA gene with NR2F1 (mirBase 20, June 2013), and more than half are conserved in target genes, the human genome (GRCh37/hg19) and mouse vertebrates [18]. In humans, genes encoding miRNAs are genome (NCBI37/mm9) builds were viewed using the UCSC located in intergenic (52%), intragenic–intronic (43%), or genome browser [27] (http://genome.ucsc.edu) with the NR2F1 intragenic-exonic (5%) regions with both sense and antisense binding site custom track supplied by Montemayor et al. [12]. orientations [19]. This large family of 21-23 nucleotide single- The miRNA track was also viewed to identify every miRNA stranded non-coding RNAs negatively regulate gene mentioned by Wang et al. to be expressed in the developing expression at the post-transcriptional level by binding the mouse inner ear [22]. Other miRNAs were identified by 5’UTR, coding sequences, or 3’UTR. They have the ability to reviewing the literature involving ontogenesis or either promote degradation or suppress the translation of their NR2F1[21,22,28–30]. To find all possible gene targets for each target mRNA, depending on the complementarity with which miRNA, we utilized multiple database prediction tools: they bind [15,16]. These small non-coding RNAs are first miRBase (http://www.mirbase.org) [31], MicroCosm (http:// formed as pri-miRNAs which are capped, adenylated, spliced, www.ebi.ac.uk/enright-srv/microcosm/htdocs/targets/v5), and later cleaved by Drosha and Pasha endonucleases TargetScan 6.0 (http://www.targetscan.org), microRNA.org (RNase-III type) resulting in precursor miRNAs (pre-miRNA) (http://www.microrna.org/microrna/home.do), and PITA (http:// [16]. Pre-miRNAs are then processed by Dicer to form short genie.weizmann.ac.il/pubs/mir07/mir07_dyn_data.html) [32]. duplex RNAs, which are incorporated into a miRNA-induced The predicted genes from the different databases were silencing complex (miRISC), in order to interact with the target combined and redundant genes removed. The final lists were mRNA. compared with the putative NR2F1 targets identified by MicroRNAs have a specific role in the developing inner ear Montemayor et al. [12]. and hearing, where many miRNA expression profiles change both temporally and spatially during embryogenesis and postnatal maturation and are conserved through evolution [20–22]. miRNA binding site predictions The significance of miRNAs in audition is exemplified by MicroRNA.org (http://www.microrna.org/microrna/home.do) mutations in the seed region of human miRNA-96 causing a was used to determine binding sites within the translated nonsyndromic progressive hearing loss in humans [21]. regions of the genes while TargetScan 6.2 (http:// Furthermore, Dicer is required for survival of inner ear hair cells www.targetscan.org) [33] and miRWalk (http://www.umm.uni- in mice, and a loss-of-function results in aberrant hair bundle heidelberg.de/apps/zmf/mirwalk/index.html) [34] were used to formation, stereocilia defects, disrupted inner ear determine binding sites in the 3’ UTR. The range of the binding morphogenesis, and impaired hearing function [23,24]. sites was manually curated. Given the apparent roles of NR2F1 and miRNAs in the ontogenesis of the inner ear, and more notably the functional Functional annotation and biological significance deficits caused when these regulators are mutated or deleted, The Database for Annotation, Visualization and Integrated we sought to discover the genetic coregulatory networks of Discovery (DAVID) [35,36] querying UniProt and the Genetic NR2F1 and miRNAs. We hypothesized that knowing if, how, Association Database was used to determine the expression and when gene regulatory pathways intersect could shed light pattern and associated neurological diseases of the 28 genes on the current incomplete knowledge of the regulatory targeted by at least 3 of the 11 miRNAs. Functional clusters mechanisms of inner ear development and functional were discovered using the Gene Functional Classification maturation [20], and perhaps extend understanding to other function in DAVID. Cancer-related genes targeted by both organ systems as well. Such mechanistic understanding is a NR2F1 and the 11 miRNAs were identified by searching the crucial step in discovering innovative ways of protecting, Sanger Institute’s cancer database. To discover genes that

PLOS ONE | www.plosone.org 2 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

affect hearing loss, the hereditary hearing loss databases Expression levels of both mRNA and miRNAs were

(hereditaryhearingloss.org), the Corey Lab at Harvard [37,38], compared using the relative CT (cycle number) method [42] and commercial hearing loss databases (SABiosciences, after normalization to the expression level of cyclophilin A. Frederick, MD) were mined and compared with the putative Primer sequences are included in Table S7 in File S1. targets of NR2F1 and the 11 miRNAs. In addition, we examined genes involved in circadian rhythm have been shown Chromatin-immunoprecipitation to be involved in many aspects of embryogenesis [39] and Cerebral cortex was harvested from C57Bl/6 mice and identified 18 genes that are targeted by at least one of the 11 processed by chromatin crosslinking and immunoprecipitation miRNAs. Finally, lists of genes involved in Notch signaling, as described in the Active Motif ChIP-IT Express Kit (Active neurogenesis, and stem cell function were obtained from Motif, Carlsbad, CA). Briefly, 150 mg of cortex was minced and SABiosciences (Frederick, MD) and filtered against the putative cross-linked with 1% formaldehyde at room temperature with targets of the 11 miRNAs. gentle shaking for 12 min. Glycine was added (final concentration of 125 mM) to quench the formaldehyde by RNA extraction and real-time polymerase chain incubating for 5 min with gentle shaking. The tissue pieces reaction were then dounce homogenized (loose clearance, 2ml, Inner ears of P0 Nr2f1–/– (–/–) and WT (+/+) littermates were K885301-002, Kontes, Fisher Scientific, Waltham, MA) in 2 ml placed in RNA Later (Ambion) after dissection, stored at 4°C of ice-cold PBS supplemented with 10 ul proteinase inhibitor overnight and then at -20°C until the genotype was determined cocktail (PIC, Active Motif, Carlsbad, CA) and 10 ul [9]. Total RNA was isolated using the RNeasy Mini Kit (Qiagen) phenylmethylsulfonyl fluoride (PMSF, 100mM, Active Motif, according to the manufacturer’s instructions and quantified Carlsbad, CA) to obtain a single-cell slurry which were pelleted using NanoDrop spectrophotometry (Wilmington, DE). at 660xg at 4°C for 5 min. Cells were resuspended in 2 ml Reverse transcription and real-time PCR of mRNAs were Lysis Buffer supplemented with PIC and PMSF, incubated for carried out using 0.25μg of RNA, nuclease-free water, 0.5μl of 30 min on ice before being dounce-homogenized again (tight Oligo(dT), 10μl 5x First-Strand buffer, 4μl of 10mM dNTP, 5μl clearance, 2ml, K885301-002, Kontes, Fisher Scientific, of 0.1M DTT, 1μl ribonuclease inhibitor (Invitrogen), and 2μl Waltham, MA) to release the nuclei, which were then pelleted SuperScript II (Invitrogen). Reverse transcription was with centrifugation for 10 min at 660xg rpm at 4°C. Chromatin performed at 42°C for 1h before heating to 70°C for 15 min fragmentation was performed in shearing buffer using a 450 then returning to 4°C. Real-time PCR was performed in Branson Sonifier (50 bursts of 30 sec pulse with 30 sec OFF, triplicates in 96-well plates in Mastercycler ep realplex 30% power output and 90% duty cycle) and the sheared (Eppendorf, Hamburg, Germany) using the following program: chromatin was collected at 20,000xg for 15 min at 4°C, after 1) 50°C for 2min, 2) 95°C for 2min, 3) 45 cycles of 95°C for which the supernatant was stored at -80°C. 15s, 60°C for 1min, and 4) melting curve ramp. Chromatin-immunoprecipitation (ChIP) was performed at 4°C Reverse transcription and real-time PCR of microRNAs were overnight with antibodies against NR2F1 (Perseus proteomics carried out using a protocol modified from Varkonyi-Gasic et al. Inc., PP-H8132, Tokyo, Japan) and histone H3K9-Ac (Cell using stem-loop primers and SYBR Green without any signaling, C5B11, Danvers MA) while control commercial miRNA detection kit [40]. All stem-loop, forward immunoprecipitations (IPs) were performed with mouse IgG and reverse primers were taken from Tang et al. [41] except for (Active Motif, Carlsbad, CA). Twenty microliters of magnetic the stem-loop and forward primers for mmu-miR-1192 which beads (Active Motif), 2 ug of antibody, and the chromatin have the sequences 5’- samples were rotated overnight at 4°C in 200 ul volumes each. ctcaactggtgtcgtggagtcggcaattcagttgagaatttggt-3’ and 5’- The next day, beads were washed twice with 800 ul ChIP acactccagctgggaaacaaacaaaca-3’. Briefly, 0.25μg of RNA for Buffer 1 and four times with 800 ul ChIP Buffer 2. The beads each sample was incubated for 5 minutes at 65°C with were then resuspended in Elution Buffer and rotated for 15 min nuclease-free water and 0.5μl of 10mM dNTP mix in a final at room temperature. Fifty microliters of Reverse-Cross-linking volume of 12.65μl and returned to ice for at least 2 minutes. A Buffer were added and the supernatant transferred to a fresh mix of 4μl 5X First-Strand buffer, 2μl 0.1M DTT, 0.1μl tube. ChIP and input samples were denatured at 95°C for 15 ribonuclease inhibitor (Invitrogen), 0.25μl SuperScript II RT min in a BioRad thermocycler before the proteins were (Invitrogen) and 1μl of 1μM stem-loop primer was added for digested with 2 μl Proteinase K (0.5ug/ul) for 1 hour at 37°C. each reaction. Reverse transcription was performed using the Genomic DNA was isolated using Qiagen PCR Kit (Qiagen, following program: 1) 16°C for 30min, 2) 45 cycles at 30°C for Valencia, CA) and used for RT-PCR as described above. 30s, 42°C for 30s, and 50°C for 1s, and 3) 85°C for 5min. For real-time PCR, 2μl of the reverse transcribed samples were Luciferase reporter assay mixed with 1μl each of the forward and reverse primers (1μM), A luciferase reporter construct containing a 3304-bp 4μl of Express SYBR Green (Invitrogen) and 12μl nuclease- fragment of the KLF9 3’ untranslated region (UTR) and an free water. The real-time PCR was performed in triplicates in empty luciferase reporter construct were purchased from 96-well plates in Mastercycler ep realplex (Eppendorf, SwitchGear Genomics (Menlo Park, CA; #S813897 and Hamburg, Germany) using the following program: 1) 95°C for #S890005, respectively). Putative miR-140 binding sites were 5min, 2) 45 cycles of 95°C for 5s, 60°C for 10s, 72°C for 1s, 3) identical with Klf9 3’UTR. miR-140 mimic and a non-targeting melting curve ramp from 65°C to 95°C at 0.1°C per second. scramble miR mimic were purchased from Life Technologies

PLOS ONE | www.plosone.org 3 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

(Grand Island, NY) and used at a final concentration of 30 nM. miR-181b, miR-183, miR-191, miR-194, miR-199b, miR-341, Transfection and luciferase assay were carried out according to and miR-1192) were selected that might participate to the manufacturer’s published protocol [43], using HEK293 cells coordinate with NR2F1 to regulate inner ear gene expression. and LipoD293 as the transfection reagent (SignaGen Laboratories, Rockville, MD). Twenty-four hours after Identification of NR2F1 gene targets co-regulated by transfection, cells were harvested and luciferase signal the 11 miRNAs assayed using LightSwitch Luciferase Assay Reagents To determine the relative potential for coregulation of target (SwitchGear Genomics) according to the manufacturer’s genes by NR2F1 and the miRNAs, we generated a list of the instructions. Luciferase signal was normalized to the total total number of genes targeted by the 11 miRNAs, as well as protein concentration of the lysates. Three independent the subset of target genes whose expression was changed in transfection experiments were performed. Statistical analysis the Nr2f1–/– mice, as quantified by microarray expression was performed using two-tailed Student’s t-test with α=0.05. profiling and/or validated by qRT-PCR [12] (Table S1 in File S1). Of the 7201 unique miRNA genes targeted by the 11 Statistics miRNAs, 107 genes are also targeted by NR2F1 and 16 genes Data are presented as mean ±SEM. Statistical analysis was were previously validated to be changed in Nr2f1–/– [12] (Figure performed using two-tailed Student’s t-test with α=0.05. 1 and Table 1). We next selected the genes targeted by at least 3 miRNAs Results for an in-depth analysis. This group of 28 genes, along with the miRNAs by which they are targeted, and the relative change in Intersection of NR2F1 targets and miRNAs expressed expression in the Nr2f1–/– [12] is shown in Table S2 in File S1. in the mouse inner ear All 28 genes are targeted by multiple miRNAs: 16 by 3 A multi-step approach was used to discover miRNAs that miRNAs, 8 by 4 miRNAs, 2 by 5 miRNAs, 1 each by 6 and 8 might participate to coregulate NR2F1 target genes. A list of miRNAs. One notable miRNA target is Nr2f1 itself that is putative NR2F1 target genes identified by Montemayor et al. targeted by three of these miRNAs (miR-17, -33 and -140), [12] was used to find positional overlap between NR2F1 target which suggests a feedback regulatory system may be possible. genes with gene loci encoding miRNAs using the UCSC Other interesting observations include that Pdgfrα, a gene genome browser. Of all the NR2F1 target genes inspected, known for its role in both development and oncogenesis only one miRNA gene intersection was found in the human [52,53], was targeted by 8 of 11 miRNA, more than any of the genome–the Klf9 gene, which contains the gene encoding other genes. Other genes with obvious roles in neurogenesis miR-1192, was previously validated to be up-regulated and and neuron maintenance are also featured on this multi-hit list, inner ear of Nr2f1–/– mice [12]. Beyond that, the most proximal such as Psen1 and Nefl [54,55]. Analysis of the pathways, miRNA genes found to intersect NR2F1 target genes were tissues, and diseases associated with each gene according to those for miR-693 and miR-1954, which were approximately the Database for Annotation, Visualization and Integrated 18kbps upstream and downstream from the Yipf3 and Ak1 Discovery (DAVID) [35,36] querying the universal protein genes, respectively. resource in UniProt [56] and the Genetic Association Database Next, we investigated recent articles that identified miRNAs (GAD) [57], revealed 16 of the 28 genes (targeted by at least 3 in the developing embryonic mouse inner ear [22–24,30,44]. Of miRNAs) are expressed in the human brain, and 18 of 24 the miRNAs expressed, miR-17 and miR-341 have NR2F1 genes have higher expression in the brain than other mouse binding sites within their coding regions, miR-140, miR-191 and tissues. Furthermore, several genes are known to cause miR-199b have NR2F1 binding sites within the proximal diseases related to neurological development and are listed in promoter region of their loci, and miR-183 and miR-181b have Table S3 in File S1. NR2F1 binding sites 8 and 15 base pairs immediately Multiple binding sites for a miRNA on a target mRNA might upstream of the transcription start sites, respectively. The indicate greater regulatory control by that miRNA, whereas miR-183 family of miRNAs is expressed and required for binding motifs shared by multiple miRNAs may indicate a lack maintenance and survival of hair cells [45,46]. There were of specificity in regulation [58]. Thus, we closely analyzed the many others that had binding sites more distant from the specific miRNA binding site in both the untranslated (UTR) and miRNA genes and were not investigated in this study. translated regions of the target mRNA. In the 3’UTR, 14 of the Additionally, miR-96, miR-140, and miR-194 were included for target genes have only one motif per miRNA, 8 have two the following reasons. miR-96 is the only miRNA demonstrated binding sites, and 6 have three or more binding sites and thus to cause hearing loss when mutated [21,28]; miR-140 is highly possibly maintain tighter regulatory control (Table S4 in File and selectively expressed in hair cells during development and S1). In the translated portion of the mRNA of target genes, targets Jag1 and RARβ, both important for hair cell and inner there were also some interesting finds as listed in Table S5 in ear development [22,30]; miR-194 targets Fgf10, which is File S1. While there was no redundancy discovered in the essential for the development of the otic vesicle [27,47,48]. 3’UTR of the target genes, 13 of the genes have binding motifs Finally, miR-33 was included because it and NR2F1 are that are compatible with more than one miRNA. involved with cholesterol metabolism which may participate in Regarding the genes that affect hearing loss, we mined hair cell development and function [29,49–51]. With these databases from the Hereditary Hearing Loss criteria a set of 11 miRNAs (miR-17, miR-33, miR-96, miR-140, (hereditaryhearingloss.org), the Serial Analysis of Gene

PLOS ONE | www.plosone.org 4 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 1. Intersection of genes targeted by both NR2F1 and select miRNAs. (A) Identification of 107 genes by intersection of the 182 NR2F1 targets genes [12] with targets of the 11 miRNAs and (B) the percentage of these genes targeted by multiple miRNAs. doi: 10.1371/journal.pone.0083358.g001

Table 1. List of 107 predicted miRNAs and NR2F1 coregulated inner ear targets.

MicroRNA Putative miRNA/NR2F1 targets mmu-miR17 Acox1, Ahnak, Ampd3, C1qa, Cidea, Cldn1, Dlk1, Hsd17b11, Kcna5, Klf9, Ldlr, Nr2f1, Pdgfra, Ppia, Sfrs2, Sqstm1, Vps25 mmu-miR-33 Adora1, Ahnak, Ampd3, Ankrd1, Casq1, Ctbp1, Cxcl12, Ddx3x, Dusp1, Eif4b, Fabp7, Hspb8, Kcna5, Mmp14, Nefl, Nr2f1, Pdgfra, Sat1, Sfrs2, Txnrd1 1810015C04Rik, Acan, Ahnak, Ak1, AK162128, Ampd3, Arg1, Cd83, Cldn1, Crym, Ddx3x, Elavl1, Foxk2, Glul, Hmgcs1, Ldlr, Man2a1, Myadm, Pdgfra, mmu-miR-96 Psen1, Sesn1, Slc25a1, Snai2, Stim1, Tbx15, Tectb, Ttr, Tyms, Zrsr1 mmu-miR-140 Acox1, C1qc, Ddx3x, Dpep1, Eln, Gjb2, Gpx3, H2afy, Klf9, Man2a1, Nr2f1, Pdgfra, Prrx2, Ptgis, Rbp1, Tectb, Timm23, Trip4, Vpreb3, Zbtb16 Acox1, Ahnak, Cd83, Chordc1, Crym, Cyr61, Ddit4, Ddx3x, Dr1, Eln, Erh, Foxk2, Hist1h2ao, Mmp14, Ms4a6b, Nbr1, Pdgfra, Pros1, Ran, Rbm8a, Sat1, mmu-miR-181b Sfn, Snai2, Sqstm1, Tcn2 mmu-miR-183 2810410M20Rik, Acan, Ak1, Cldn1, Ddx3x, Dlk1, Eif4b, Foxk2, Gjb2, Ndufb6, Nefl, Ppia, Psen1, Ran, Sfrs2, Tbx15, Wdhd1 mmu-miR-191 Dr1, Elavl1, Foxk2, Mgp, Nbr1, Sat1, Satb1, Tsen34 2410022L05Rik, Cd83, Cidea, Cldn1, Cxcl12, Cyp51, Dr1, Eln, Erh, Fkbp5, Glul, Hmgcs1, Nbr1, Pdgfra, Ran, Rasa4, Scg2, Sfrs2, Tmem106a, Txnrd1, mmu-miR-194 Vps25, Ywhab mmu-miR-199b Aaas, C1qa, Ddit4, Dr1, Eln, Fabp7, Ldlr, Ltc4s, Pdgfra, Pros1, Psen1, Relb, Sat1, Sfrs2, Stim1, Tcp11, Tyro3, Vps4a mmu-miR-341 C1qc, Crabp1, Hsd17b11, Jun, Nfkbia, Ptgis, Rbm8a Ampd3, Chordc1, Dbi, Ddit4, Ddx3x, Dhx40, Dr1, Dusp1, Elavl1, Fkbp5, Foxk2, Glul, Hmgcs1, Jun, Klf9, Ldlr, Man2a1, Nedd9, Nefl, Pdgfra, Pros1, Ptgis, mmu-miR-1192 Ran, Satb1, Scg2, Stim1, Tbx15, Tek, Tsen34, Wdhd1, Ywhab Genes in bold type were previously validated to be changed in Nr2f1–/– tissues by qRT-PCR [12]. doi: 10.1371/journal.pone.0083358.t001

Expression (SAGE) and the literature [37,38], and commercial Functional clustering of the genes co-targeted by the hearing loss databases and found four genes to be targeted by 11 miRNAs and NR2F1 both NR2F1 and miRNAs: Crym and Snai2 were both targeted Using the Gene Ontology (GO) Gene Functional by miR-96 and miR-181b; Gjb2 was targeted by miR-140 and Classification analysis and the DAVID database [35,36], we miR-183 (Table S5 in File S1). In addition, 69 other genes discovered 9 functional clusters which includes all the 28 co- implicated in hearing loss were targeted solely by the miRNAs targeted genes (Table S2 in File S1) and contains 66 of the but not NR2F1 (not shown). 107 genes targeted by at least 1 of the 11 miRNAs and NR2F1

PLOS ONE | www.plosone.org 5 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 2. MicroRNA expression analyses in Nr2f1–/– tissues. (A) The expression of select miRNAs in the Nr2f1–/– inner ear showing miR-140 and miR-181a were significantly down-regulated by 4.5-fold (P=0.004) and 1.7-fold (P=0.046), respectively, compared to the WT. The levels of miR-181b, -181c and -191 were also decreased but did not reach significance (P=0.3, 0.4, 0.1, respectively). (B) The expression of miR-140 and miR-181a were unchanged in the cerebral cortex tissues from the same animals. N=3 in each group. Significance was assessed by a two-tailed t-test for independent comparison between Nr2f1–/– and WT values for each miRNA with α=0.05. doi: 10.1371/journal.pone.0083358.g002

(Table S6 in File S1). Based on GO enrichment score, Cluster in the Klf9 gene which is up-regulated in the inner ear of 1 genes (Crabp1, Fabp7, Rbp1) involve cellular transport of Nr2f1–/– mouse [12] and is targeted by miR-140 (Table 1). By fatty acids and hormones in metabolic process, including analyzing these 6 miRNAs using RT-PCR, we found that –/– forebrain development, which is defective in Nr2f1 . Cluster 2 miR-140 and miR-181a are significantly down-regulated in the genes (Eif4b, Elavl1, Rbm8a, Sfrs2, Zrsr1) involve RNA binding Nr2f1–/– inner ear by 4.5-fold (P=0.004) and 1.7-fold (P=0.046) and metabolic processes and post-translational regulation of compared with wildtype (WT), respectively, while the other 4 biological processes. The 10 genes in clusters 3-5 involve miRNAs were not significantly down-regulated (Figure 2A). steroid, lipid and cholesterol biosynthesis, cation transport and Because Nr2f1–/– mice have various neural defects [5,6,9–11], chromatin assembly and organization, which are key processes during development. Cluster 6 has 20 genes that explicitly we went a step further to quantify the levels of miR-140, -181a, –/– involve transcription, and include Nr2f1 and Klf9. Clusters 7 and -1192 in the cerebral cortex of the Nr2f1 but found that through 9 have lower enrichment scores but involves skeletal their levels were not significantly changed (Figure 2B). This development, nucleotide binding, cell adhesion, tube suggests a tissue-specific role for miR-140 and miR-181a in development and morphogenesis, the endomembrane system the Nr2f1–/– inner ear. and energy regulation, many functions of which are defective in Nr2f1–/– [8,9,59]. NR2F1 directly regulates miR-140 Since the level of miR-140 was the most down-regulated in –/– Validation of miRNA as targets using Nr2f1 tissues the absence of NR2F1 (Figure 2A), we investigated whether In order to validate this bioinformatics approach, we pursued NR2F1 might directly regulate miR-140 expression. We –/– several miRNAs by determining their levels in the Nr2f1 mice identified putative NR2F1 binding sites immediately upstream using qRT-PCR. Of note, miR-140, -181b and -191 each have of the miR-140 locus [12] (Figure 3A) and performed a putative NR2F1 binding site close to the proximal portion of chromatin-immunoprecipitation (ChIP) for NR2F1 binding their genes and since miR-181b is found in a family cluster we enrichment followed by qRT-PCR analysis. Compared to a also included miR-181a and -181c, which may be coregulated non-specific IgG control, there is a significant NR2F1 binding and share similar gene targets. MicroRNA-140 has 4 co- targeted genes with NR2F1 amongst the validated expression enrichment (P<0.05) at the miR-140 locus (Figure 3B) changes in the Nr2f1–/– mice and is predicted to target NR2F1 coincident with enrichment of the open chromatin marker itself (Table 1 & S1 in File S1). miR-1192 is the only miR locus acetylated H3K9, suggesting NR2F1 might directly regulate that is located within any of the NR2F1 target genes; it is found miR-140 transcription.

PLOS ONE | www.plosone.org 6 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 3. NR2F1 directly regulates miR-140. (A) Putative NR2F1 binding sites (red boxes) immediately upstream of the miR-140 locus [12]. Thick arrows define regions of qRT-PCR amplification. (B) Chromatin-immunoprecipitation (ChIP) and qRT-PCR analyses reveal a significant enrichment of the miR-140 locus with NR2F1 pull-down. N=5-6. *P<0.05, $P=0.1 vs IgG control. doi: 10.1371/journal.pone.0083358.g003

Tissue-Dependent Co-regulation of Klf9 by NR2F1 and further validating the bioinformatically predicted coregulatory miR-140 network involving NR2F1, miR-140 and Klf9. We found it intriguing that miR-1192 expression was not altered whereas the gene Klf9, which harbors the miR-1192 Discussion locus, was up-regulated in the Nr2f1–/– [12], suggesting that Klf9 is regulated independently in the inner ear. Klf9 is one of Since their discovery approximately 25 years ago [60], there the 28 genes putatively targeted by both NR2F1 and at least 3 has been great interest in orphan nuclear receptors and their of the 11 miRNAs in this study (Figure 1 & Table S2 in File S1). putative ligands. This is based on the knowledge that nuclear In particular, Klf9 has multiple miR-140 binding sites: 3 sites in receptors and cofactors play crucial roles in many physiological its 3’ UTR and 4 sites in its translated region (Tables S4 & S5 processes during development, reproduction, metabolism and in File S1). To determine whether miR-140 regulates Klf9 aging [3,61–63]. Orphan nuclear receptors have been directly, we tested if a miR-140 mimic would target and affect discovered to be key components in understanding the the expression of a luciferase gene construct containing the regulatory milieu at the genomic scale as they translate KLF9 3’UTR sequences (Figure 4A). We quantified Luciferase cellular, paracrine, neuronal and environmental stimuli into activity from the luciferase-KLF9 3'UTR reporter and found that epigenetic and gene expression signals [64–67]. These signals miR-140 mimic reduced the Luciferase signal by more than in turn regulate cellular and organismal homeostasis, and 90% compared to a scrambled mimic (P<0.001, Figure 4B). dysregulation of this process can lead to diseases including Next, we determined whether NR2F1 regulates Klf9 directly by cancer (Tables S3 & S6 in File S1) [68]. performing a ChIP followed by qRT-PCR using primers that The orphan nuclear receptor NR2F1, located on amplify the Klf9 first exon (Figure 4C). Impressively, there was chromosome 5q14, and its homolog NR2F2, are the most a 500-fold enrichment of NR2F1 binding at the Klf9 locus as evolutionarily conserved nuclear receptors across all species compared to a non-specific IgG control (P<0.05) (Figure 4D). [69]. NR2F1 is implicated in neurogenesis, cell fate Since miR-140 is down-regulated in the absence of NR2F1 determination, patterning, and regulation of differentiation in the only in the inner ear and not the cerebral cortex (Figure 2) and rodent inner ear and cerebral cortex [7–10]. We report here a is a direct target of NR2F1 (Figure 3), we tested if Klf9 was proof-of-concept novel coregulatory network involving the regulated in a tissue-dependent manner as well. Indeed, using orphan nuclear receptor NR2F1 and microRNA-140 with qRT-PCR, we confirmed that Klf9 is up-regulated almost 3-fold experimental validation of the effects these regulators have on in the inner ear (P<0.01 vs WT) but was unchanged in the each other as well as a mutual target, the Klf9 gene. This study cerebral cortex of Nr2f1–/– mice (Figure 5). Consistent with the exemplifies the research approach of using systematic luciferase-Klf9-3’UTR activity that demonstrated a strong bioinformatic data mining to uncover coregulatory interactions repression of Klf9 by the miR-140 mimic (Figure 4B), these followed by quantitative experimental validations of regulated data suggest that miR-140 may be a determining factor in targets at the transcription, post-transcription and genomic regulating Klf9 expression that is represented in the levels. The validation included qRT-PCR, luciferase-3’UTR coregulatory network of NR2F1, miR-140 and Klf9 (Figure 6). reporter and ChIP-qRT-PCR assays with tissues from wild type Together, these experiments provide molecular evidence that and mutant mice. This approach accomplished several Klf9 is directly coregulated by both miR-140 and NR2F1, objectives: 1) discovery of a mechanism by which NR2F1

PLOS ONE | www.plosone.org 7 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 4. Both miR-140 and NR2F1 directly regulate Klf9 expression. (A) Putative miR-140 binding sites (red boxes) on KLF9-3’UTR. (B) Luciferase expression from a KLF9 3'UTR reporter is significantly reduced in the presence of the miR-140 mimic while Luc expression from the empty reporter is unchanged. N=3. ***P<0.001. (B) Schematic of the putative NR2F1 binding site (red box) just downstream of the first exon of Klf9 and the relative position of the qRT-PCR primers (Thick arrows). (D) Log-scale presentation of qRT-PCR results following chromatin-immunoprecipitation (ChIP) demonstrates significant enrichment of NR2F1 at the Klf9 locus. N=5. *P<0.05, $P=0.08 vs IgG control. doi: 10.1371/journal.pone.0083358.g004 regulates direct target gene expressions; 2) identification of of the vestibular apparati by birth [22]. These 3 separate specific miRNAs, such as miR-140, that co-regulate and alter expression patterns suggest miR-140 might have stage- NR2F1 target gene expressions, such as Klf9; 3) delineation of specific roles during inner ear development. Indeed, SOX9 a feedback coregulation that occurs between NR2F1 and directly activates miR-140 [74] and has a differentially and miR-140; and 4) identification of a tissue-dependent overlapping expression as SOX2, and is consistent with the coregulatory network involving NR2F1, miR-140 and Klf9. miR-140 expression, during cochlear development [75]. This The miR-140 gene, located on chromosome 16q22, is most pattern is also highly consistent with that of Nr2f1 expression notable for its role in cartilage development and maintenance during inner ear development [7], supporting the possibility of [70,71]. It also has a suppressive action in liver tumorigenesis direct regulation of miR-140 by NR2F1. Indeed, we found [72] and induces pluripotent cells to differentiate into miR-140 transcripts were significantly down-regulated by 4.5- adipocytes [73]. Within the inner ear, miR-140 exhibits fold in Nr2f1–/– newborn inner ears but remained unchanged in differential expression across developmental stages and post- the cerebral cortex (Figure 2), suggesting NR2F1 acts as a natal maturation [22]. Expression is initially limited to the medial tissue-specific transactivator of miR-140. On the other hand, otocyst where cells are proliferating. As development the Nr2f1 gene itself is a putative target of miR-140 (Table S2 progresses, miR-140 is then expressed in the early sensory in File S1), a relationship that may represent a negative epithelium and subsequently in the inner and outer hair cells of feedback loop between NR2F1 and miR-140 as proposed in the organ of Corti. Lastly, miR-140 is also detected in hair cells Figure 6. Since this finding purports a regulatory feedback loop,

PLOS ONE | www.plosone.org 8 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 5. Klf9 is a downstream inner ear target of NR2F1. Klf9 expression is significantly up-regulated in Nr2f1–/– (KO) inner ear but not in the cerebral cortex, as determined by qRT-PCR. N=3. **P<0.01. doi: 10.1371/journal.pone.0083358.g005 we investigated closely the relationship between NR2F1, Further empiric evidence exists of coregulation between miR-140, and a mutual target, Klf9. miR-140, NR2F1, and Klf9. Klf9 is regulated by thyroid Klf9, located on 9q21, acts as a circadian output factor [76], hormone and its nuclear receptor TRβ, which are essential for has a role in the regulation of adipogenesis (similar to miR-140) normal development and functional maturation of hearing [77], and is up-regulated in a cortisol and differentiation-state- [85–87]. NR2F1, in turn, is a negative regulator of thyroid dependent manner. Furthermore, gain- and loss-of-function hormone receptor binding and thereby suppresses the results in growth arrest and proliferative effects, respectively, in activation of its target genes (e.g. Klf9) [86,88]. Additionally, keratinocytes and cancer stem cells [76]. Since all three genes KLF9 directly suppresses Notch1 expression and its (Nr2f1, miR-140 and Klf9) are expressed in the developing downstream signaling which promotes differentiation and inhibits glioblastoma neurosphere growth [89]. Indeed, NR2F1 inner ear and adult brain (including the cerebral cortex) during inhibits Notch signaling leading to development and neurogenesis, there are other tissues where miR-140, NR2F1, differentiation of extra hair cells and supporting cells in the and Klf9 functionally interact, but the mechanisms of their inner ear [8,20]. Cumulatively, these expression patterns and interactions have not yet been delineated. Indeed, both KLF9 described actions of NR2F1, Klf9 and miR-140 are consistent and NR2F1 have a role in lipid processing [76,77], and with the co-regulatory network proposed (Figure 6). Our miR-140 and Klf9 are both regulated in the brain by estrogen- findings indicate that NR2F1 directly regulates Klf9 as well as receptor modulation [7,22,78]. In the inner ear, estrogen acts to miR-140, while miR-140 has a direct inhibitory effect on Klf9 respond to noise-induced damage, and loss of ERβ causes transcripts. The indirect regulatory effect of NR2F1 by means inner ear defects and hearing loss [79–81]. Klf9 opposes of miR-140 might have an additive repressive effect than solely estrogen hormonal action by repressing estrogen receptor by the direct NR2F1targeting of Klf9. (ERα) expression and activity [82–84]. ERα is also known to In addition to this focused regulation of Klf9 by a regulatory down-regulate miR-140 in breast cancer stem cells suggesting network involving miR-140 and NR2F1, our bioinformatics miR-140 is downstream of estrogen action [83]. The search revealed at least 20 other common targets shared by combination of these findings may indicate that the function of miR-140 and NR2F1, four of which have been validated to mir-140 and Klf9 is important for estrogen action in the inner have a change of expression in the Nr2f1–/– mice [12] (Table ear. S1 in File S1). Klf9 is only one of these genes, which according

PLOS ONE | www.plosone.org 9 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

Figure 6. Models of transcription factor-target gene feedback and coregulatory relationships. (A) Generic model showing feedback loops and coregulation between transcription factors (TFs), miRNAs, and genes predicted to be involved in hearing, cancer, and development based on bioinformatic analyses. (B) Model of a validated coregulatory network involving NR2F1, miR-140, and Klf9. Arrow denotes either positive or negative regulation. doi: 10.1371/journal.pone.0083358.g006 to this model is transactivated at the transcriptional level by sites on some of the genes. This implicates coordination NR2F1 but repressed at the post-transcriptional level by between the different miRNAs and may allow for a greater miR-140 in order to achieve a balanced expression level. The regulation of target genes. Also, depending on how many of the remaining common targets are also potential miR-140 and regulatory miRNAs are active, this could also allow fine-tuning NR2F1 co-regulated genes. of the regulation. Second, the same miRNAs may have multiple Although we did not study them in depth, our bioinformatics binding sites. This suggests a greater or more effective search revealed a number of other possible regulatory regulatory repression and may allow for fine-tuning depending networks. There were 107 putative gene targets identified that on the accessibility of the different sites under varying were shared by NR2F1 and the 11 microRNA that we conditions [58]. In all, these findings point to the complexity of analyzed. Similar to miR-140, miR-17 and miR-33 target the interplay between these miRNAs in achieving a balanced NR2F1 and could therefore be involved in regulatory feedback regulation of the genes important for development [15]. loops with NR2F1. The majority of the genes (59%) were Furthermore, these coregulatory networks may have clinical targeted by multiple miRNA, which may indicate a higher level significance. The genes targeted by both NR2F1 and the 11 of control over expression (Figure 1B). Pdgfra, a mutation of miRNAs that we identified bioinformatically have previously which can cause neural tube defects (Table S3 in File S1) was been implicated in hearing loss (Nr2f1, Crym, Snai2, Gjb2) targeted by NR2F1 and 8 of the miRNAs. Additionally, the [11,90–94] and cancer (Pdgfrα, Eln, Jun) [52,95–98], as well as miRNAs themselves may act in concert with each other in inner ear developmental regulators (Nr2f1, Notch1, Jag 1, regulating genes important for ontogenesis and development in Nptx1, Runx1) [7,8,22,99], as listed in Tables S4 & S5 in File general. Using TargetScan 6.2 and microRNA.org to curate the S1. Nr2f1–/– mice have inner ear developmental defects binding sites on the 28 genes targeted by at least 3 of the 11 [7,8,22,99]. Furthermore, miR-140 is selectively expressed in miRNAs and NR2F1 (Tables S2, S4 & S5 in File S1), we inner ear hair cells during development [22,30] and is discovered two interesting trends. First, different miRNAs (or demonstrated to regulate targets such as PDGF [100]. It is not even the same miRNA in some cases) may bind to overlapping surprising that there is a tissue-dependent difference in

PLOS ONE | www.plosone.org 10 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

regulation of Klf9 in the inner ear compared to the cortex of Table S2. Genes targeted by at least 3 of the 11 selected Nr2f1–/– mice given the site-specific regulation of miR-140 and miRNAs. NR2F1. Nor is it unexpected that defects in the cochlear duct, Table S3. Representative list of genes targeted by multiple organ of Corti, and hair cells can occur in animals with impaired miRNAs and their associated neurological and other diseases NR2F1 or miRNA function [7,8,11,16,21,27]. These defects based on the Genetic Association Database. help to underscore the importance of NR2F1 and microRNAs Table S4. List of genes targeted by at least 3 of the 11 select like miR-140, both individually and as coregulatory partners, miRNAs and the respective miRNA binding sites in their mRNA and give clinical importance to the networked described here. 3’ untranslated region based on TargetScan 6.2. In conclusion, we have demonstrated that NR2F1 and the Table S5. Genes targeted by at least 3 of 11 selected miRNAs 11-selected miRNA may play an important or indispensable and the respective miRNA binding sites based on miRanda role in ontogenesis and development in general. As a proof-of- analysis from microRNA.org. principle, we experimentally demonstrated the coregulatory Table S6. Functional classification of the genes targeted by mechanisms and feedback loops between NR2F1, miR-140, miRNAs and NR2F1. and Klf9. We also uncovered several other potential pathways Table S7. List of primers for real-time RT-PCR. between NR2F1, miRNAs, and genes important for (DOCX) development and disease. These targets and pathways require further studies and validation in order to advance our Acknowledgements understanding of ontogenesis and provide novel strategies for protecting or regenerating hair cells and sensory neurons in D.Y.C. is a trainee in the Baylor College of Medicine Medical patients with hearing loss. Scientist Training Program (MSTP).

Supporting Information Author Contributions

Conceived and designed the experiments: DYC DWC NL FAP. File S1. Supporting Information File containing Tables S1 Performed the experiments: DYC DWC FRR. Analyzed the to S7: data: DYC DWC. Contributed reagents/materials/analysis tools: Table S1. Number of predicted miRNAs and NR2F1 inner ear NL FAP. Wrote the manuscript: DYC DWC NL FAP. targets.

References

1. Rosenfeld MG, Glass CK (2001) Coregulator codes of transcriptional 12. Montemayor C, Montemayor Oa, Ridgeway A, Lin F, Wheeler et al Da. regulation by nuclear receptors. J Biol Chem 276: 36865–36868. doi: (2010) Genome-wide analysis of binding sites and direct target genes 10.1074/jbc.R100041200. PubMed: 11459854. of the orphan nuclear receptor NR2F1/COUP-TFI. PLOS ONE 5: 2. Tsai SY, Tsai MJ (1997) Chick ovalbumin upstream promoter- e8910. doi:10.1371/journal.pone.0008910. PubMed: 20111703. transcription factors (COUP-TFs): coming of age. Endocr Rev 18: 229– 13. Wasserman WW, Sandelin A (2004) Applied bioinformatics for the 240. doi:10.1210/er.18.2.229. PubMed: 9101138. identification of regulatory elements. Nat Rev Genet 5: 276–287. doi: 3. Giguère V (1999) Orphan nuclear receptors: from gene to function. 10.1038/nrg1315. PubMed: 15131651. Endocr Rev 20: 689–725. doi:10.1210/er.20.5.689. PubMed: 14. Pandey DP, Picard D (2010) Multidirectional interplay between nuclear 10529899. receptors and microRNAs. Curr Opin Pharmacol 10: 637–642. doi: 4. Olefsky JM (2001) Nuclear receptor minireview series. J Biol Chem 10.1016/j.coph.2010.08.009. PubMed: 20829111. 276: 36863–36864. doi:10.1074/jbc.R100047200. PubMed: 11459855. 15. He L, Hannon GJ (2004) MicroRNAs: small RNAs with a big role in 5. Armentano M, Filosa A, Andolfi G, Studer M (2006) COUP-TFI is gene regulation. Nat Rev Genet 5: 522–531. doi:10.1038/nrg1379. required for the formation of commissural projections in the forebrain by PubMed: 15211354. regulating axonal growth. Development 133: 4151–4162. doi:10.1242/ 16. Esquela-Kerscher A, Slack FJ (2006) Oncomirs - microRNAs with a dev.02600. PubMed: 17021036. role in cancer. Nat Rev Cancer 6: 259–269. doi:10.1038/nrc1840. 6. Armentano M, Chou S-J, Tomassy GS, Leingärtner A, O’Leary DDM et PubMed: 16557279. al. (2007) COUP-TFI regulates the balance of cortical patterning 17. Subramanyam D, Blelloch R (2011) From microRNAs to targets: between frontal/motor and sensory areas. Nat Neurosci 10: 1277– pathway discovery in cell fate transitions. Curr Opin Genet Dev 21: 1286. doi:10.1038/nn1958. PubMed: 17828260. 498–503. doi:10.1016/j.gde.2011.04.011. PubMed: 21636265. 7. Tang LS, Alger HM, Lin F, Pereira F (2005) Dynamic expression of 18. Lagos-Quintana M, Rauhut R, Meyer J, Borkhardt A, Tuschl T (2003) COUP-TFI and COUP-TFII during development and functional New microRNAs from mouse and human. RNA 9: 175–179. doi: maturation of the mouse inner ear. Gene Expr Patterns 5: 587–592. doi:10.1016/j.modgep.2005.03.012. PubMed: 15907456. 10.1261/rna.2146903. PubMed: 12554859. 8. Tang LS, Alger HM, Pereira F a (2006) COUP-TFI controls Notch 19. Hinske LCG, Galante PAF, Kuo WP, Ohno-Machado L (2010) A regulation of hair cell and support cell differentiation. Development 133: potential role for intragenic miRNAs on their hosts’ interactome. BMC 3683–3693. doi:10.1242/dev.02536. PubMed: 16914494. Genomics 11: 533. doi:10.1186/1471-2164-11-533. PubMed: 9. Qiu Y, Pereira Fa, DeMayo FJ, Lydon JP, Tsai SY et al. (1997) Null 20920310. mutation of mCOUP-TFI results in defects in morphogenesis of the 20. Kopecky B, Fritzsch B (2011) Regeneration of Hair Cells: Making glossopharyngeal ganglion, axonal projection, and arborization. Genes Sense of All the Noise. Pharmaceuticals (Basel) 4: 848–879. doi: Dev 11: 1925–1937. doi:10.1101/gad.11.15.1925. PubMed: 9271116. 10.3390/ph4060848. PubMed: 21966254. 10. Zhou C, Qiu Y, Pereira Fa, Crair MC, Tsai SY et al. (1999) The nuclear 21. Mencía A, Modamio-Høybjør S, Redshaw N, Morín M, Mayo-Merino F orphan receptor COUP-TFI is required for differentiation of subplate et al. (2009) Mutations in the seed region of human miR-96 are neurons and guidance of thalamocortical axons. Neuron 24: 847–859. responsible for nonsyndromic progressive hearing loss. Nat Genet 41: doi:10.1016/S0896-6273(00)81032-6. PubMed: 10624948. 609–613. doi:10.1038/ng.355. PubMed: 19363479. 11. Brown KK, Alkuraya FS, Matos M, Robertson RL, Kimonis VE, et al. 22. Wang X-R, Zhang X-M, Zhen J, Zhang P-X, Xu G et al. (2010) (2009) NR2F1 deletion in a patient with a de novo paracentric MicroRNA expression in the embryonic mouse inner ear. Neuroreport inversion, inv(5)(q15q33.2), and syndromic deafness. American journal 21: 611–617. doi:10.1097/WNR.0b013e328338864b. PubMed: of medical genetics Part A 149A: 931–938. doi:10.1002/ajmg.a.32764. 20467336.

PLOS ONE | www.plosone.org 11 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

23. Soukup Ga, Fritzsch B, Pierce ML, Weston MD, Jahan I et al. (2009) organ-specific miRNA cluster. J Biol Chem 282: 25053–25066. doi: Residual microRNA expression dictates the extent of inner ear 10.1074/jbc.M700501200. PubMed: 17597072. development in conditional Dicer knockout mice. Dev Biol 328: 328– 45. Pierce ML, Weston MD, Fritzsch B, Gabel HW, Ruvkun G et al. (n.d.) 341. doi:10.1016/j.ydbio.2009.01.037. PubMed: 19389351. MicroRNA-183 family conservation and ciliated neurosensory organ 24. Friedman LM, Dror A, Mor E, Tenne T, Toren G, et al. (2009) expression. Evol Dev 10: 106–113. doi:10.1111/j.1525-142X. MicroRNAs are essential for development and function of inner ear hair 2007.00217.x. PubMed: 18184361. cells in vertebrates. Proceedings of the National Academy of Sciences 46. Weston MD, Pierce ML, Jensen-Smith HC, Fritzsch B, Rocha-Sanchez of the United States of America 106: 7915–7920. doi:10.1073/pnas. S et al. (2011) MicroRNA-183 family expression in hair cell 0812446106. development and requirement of microRNAs for hair cell maintenance 25. Tang LS, Montemayor C, Pereira FA (2006) Sensorineural hearing and survival. Dev Dyn 240: 808–819. doi:10.1002/dvdy.22591. loss: potential therapies and gene targets for drug development. PubMed: 21360794. IUBMB Life 58: 525–530. doi:10.1080/15216540600913258. PubMed: 47. Pauley S, Wright TJ, Pirvola U, Ornitz D, Beisel K et al. (2003) 17002980. Expression and function of FGF10 in mammalian inner ear 26. Rudnicki A, Avraham KB (2012) microRNAs: the art of silencing in the development. Dev Dyn 227: 203–215. doi:10.1002/dvdy.10297. ear. EMBO molecular medicine 4: 849–859. doi:10.1002/emmm. PubMed: 12761848. 201100922. 48. Schimmang T (2007) Expression and functions of FGF ligands during 27. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH et al. (2002) early otic development. Int J Dev Biol 51: 473–481. doi:10.1387/ijdb. The Human Genome Browser at UCSC. Genome Res 12: 996–1006. 072334ts. PubMed: 17891710. doi:10.1101/gr.229102. PubMed: 12045153. 49. Levic S, Yamoah EN (2011) Plasticity in membrane cholesterol 28. Lewis Ma, Quint E, Glazier AM, Fuchs H, De Angelis MH et al. (2009) contributes toward electrical maturation of hearing. J Biol Chem 286: An ENU-induced mutation of miR-96 associated with progressive 5768–5773. doi:10.1074/jbc.M110.186486. PubMed: 21163952. hearing loss in mice. Nat Genet 41: 614–618. doi:10.1038/ng.369. 50. Rajagopalan L, Greeson JN, Xia A, Liu H, Sturm A et al. (2007) Tuning PubMed: 19363478. of the outer hair cell motor by membrane cholesterol. J Biol Chem 282: 29. Najafi-Shoushtari SH, Kristo F, Li Y, Shioda T, Cohen DE et al. (2010) 36659–36670. doi:10.1074/jbc.M705078200. PubMed: 17933870. MicroRNA-33 and the SREBP host genes cooperate to control 51. Purcell EK, Liu L, Thomas PV, Duncan RK (2011) Cholesterol cholesterol homeostasis. Science 328: 1566–1569. doi:10.1126/ influences voltage-gated calcium channels and BK-type potassium science.1189123. PubMed: 20466882. channels in auditory hair cells. PLOS ONE 6: e26289. doi:10.1371/ 30. Dogini DB, Ribeiro P, Rocha C, Pereira TC, Lopes-Cendes I (2008) journal.pone.0026289. PubMed: 22046269. MicroRNA expression profile in murine central nervous system 52. Andrae J, Gallini R, Betsholtz C (2008) Role of platelet-derived growth development. Journal of molecular neuroscience : MN 35: 331–337. factors in physiology and medicine. Genes Dev 22: 1276–1312. doi: doi:10.1007/s12031-008-9068-4. 10.1101/gad.1653708. PubMed: 18483217. 31. Kozomara A, Griffiths-Jones S (2011) miRBase: integrating microRNA 53. Chen H, Gu X, Liu Y, Wang J, Wirt SE et al. (2011) PDGF signalling annotation and deep-sequencing data. Nucleic Acids Res 39: D152– controls age-dependent proliferation in pancreatic β-cells. Nature, 478: D157. doi:10.1093/nar/gkq1027. PubMed: 21037258. 349–55. doi:10.1038/nature10502. PubMed: 21993628. 32. Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E (2007) The role of 54. Bekris LM, Yu C-E, Bird TD, Tsuang DW (2010) Genetics of Alzheimer site accessibility in microRNA target recognition. Nat Genet 39: 1278– disease. J Geriatr Psychiatry Neurol 23: 213–227. doi: 1284. doi:10.1038/ng2135. PubMed: 17893677. 10.1177/0891988710383571. PubMed: 21045163. 33. Friedman RC, Farh KK-H, Burge CB, Bartel DP (2009) Most 55. Houlden H, Reilly MM (2006) Molecular genetics of autosomal- mammalian mRNAs are conserved targets of microRNAs. Genome dominant demyelinating Charcot-Marie-Tooth disease. Neuromolecular Res 19: 92–105. doi:10.1101/gr.082701.108. PubMed: 18955434. Med 8: 43–62. doi:10.1385/NMM:8:1:243. PubMed: 16775366. 34. Dweep H, Sticht C, Pandey P, Gretz N (2011) miRWalk--database: 56. The UniProt Consortium (2011) Ongoing and future developments at prediction of possible miRNA binding sites by “walking” the genes of the Universal Protein. Resour - Nucleic Acids Research 39: D214– three genomes. J Biomed Inform 44: 839–847. doi:10.1016/j.jbi. D219. doi:10.1093/nar/gkq1020. 2011.05.002. PubMed: 21605702. 57. Genetic Association Database (2011). Available:http:// 35. Huang DW, Sherman BT, Lempicki R (2009) Bioinformatics enrichment geneticassociationdb.nih.gov. tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 37: 1–13. doi:10.1093/nar/gkn923. PubMed: 58. Valencia-Sanchez MA, Liu J, Hannon GJ, Parker R (2006) Control of 19033363. translation and mRNA degradation by miRNAs and siRNAs. Genes 36. Huang DW, Sherman BT, Lempicki R (2009) Systematic and Dev 20: 515–524. doi:10.1101/gad.1399806. PubMed: 16510870. integrative analysis of large gene lists using DAVID bioinformatics 59. Tripodi M, Filosa A, Armentano M, Studer M (2004) The COUP-TF resources. Nat Protoc 4: 44–57. doi:10.1038/nprot.2008.211. PubMed: nuclear receptors regulate cell migration in the mammalian basal 19131956. forebrain. Development 131: 6119–6129. doi:10.1242/dev.01530. 37. Abe S, Koyama K, Usami S, Nakamura Y (2003) Construction and PubMed: 15548577. characterization of a vestibular-specific cDNA library using T7-based 60. Giguère V, Yang N, Segui P, Evans RM (1988) Identification of a new RNA amplification. J Hum Genet 48: 142–149. doi:10.1007/ class of steroid hormone receptors. Nature 331: 91–94. doi: s100380300022. PubMed: 12624726. 10.1038/331091a0. PubMed: 3267207. 38. Chen Z-Y, Corey DP (2002) An inner ear gene expression database. 61. McKenna NJ, O’Malley BW (2002) Combinatorial control of gene Journal of the Association for Research in Otolaryngology : JARO 3: expression by nuclear receptors and coregulators. Cell 108: 465–474. 140–148. doi:10.1007/s101620020029. PubMed: 12162364. doi:10.1016/S0092-8674(02)00641-4. PubMed: 11909518. 39. Seron-Ferre M, Valenzuela GJ, Torres-Farfan C (2007) Circadian 62. Cho H, Zhao X, Hatori M, Yu RT, Barish GD et al. (2012) Regulation of clocks during embryonic and fetal development. Birth Defects Res C circadian behaviour and metabolism by REV-ERB-α and REV-ERB-β. Embryo Today 81: 204–214. doi:10.1002/bdrc.20101. PubMed: Nature 485: 123–127. doi:10.1038/nature11048. PubMed: 22460952. 17963275. 63. Foulds CE, Feng Q, Ding C, Bailey S, Hunsaker TL et al. (2013) 40. Varkonyi-Gasic E, Wu R, Wood M, Walton EF, Hellens RP (2007) Proteomic analysis of coregulators bound to ERα on DNA and Protocol: a highly sensitive RT-PCR method for detection and nucleosomes reveals coregulator dynamics. Mol Cell 51: 185–199. doi: quantification of microRNAs. Plant Methods 3: 12. doi: 10.1016/j.molcel.2013.06.007. PubMed: 23850489. 10.1186/1746-4811-3-12. PubMed: 17931426. 64. Yang X, Downes M, Yu RT, Bookout AL, He W et al. (2006) Nuclear 41. Tang F, Hajkova P, Barton SC, Lao K, Surani MA (2006) MicroRNA receptor expression links the circadian clock to metabolism. Cell 126: expression profiling of single whole embryonic stem cells. Nucleic Acids 801–810. doi:10.1016/j.cell.2006.06.050. PubMed: 16923398. Res 34: e9. doi:10.1093/nar/gnj009. PubMed: 16434699. 65. Ahmadian M, Suh JM, Hah N, Liddle C, Atkins AR et al. (2013) PPARγ 42. Schmittgen TD, Livak KJ (2008) Analyzing real-time PCR data by the signaling and metabolism: the good, the bad and the future. Nat Med comparative CT method. Nat Protoc 3: 1101–1108. doi:10.1038/nprot. 19: 557–566. doi:10.1038/nm.3159. PubMed: 23652116. 2008.73. PubMed: 18546601. 66. Deblois G, St-Pierre J, Giguère V (2013) The PGC-1/ERR signaling 43. Aldred SF, Collins P, Trinklein N (2011) Identifying targets of human axis in cancer. Oncogene 32: 3483–3490. doi:10.1038/onc.2012.529. micrornas with the LightSwitch Luciferase Assay System using 3’UTR- PubMed: 23208510. reporter constructs and a microRNA mimic in adherent cells. Journal of 67. Rada-Iglesias A, Bajpai R, Prescott S, Brugmann SA, Swigut T et al. Visualized Experiments : JoVE: 3–7. doi:10.3791/3343. (2012) Epigenomic annotation of enhancers predicts transcriptional 44. Xu S, Witmer PD, Lumayag S, Kovacs B, Valle D (2007) MicroRNA regulators of human neural crest. Cell Stem Cell 11: 633–648. doi: (miRNA) transcriptome of mouse retina and identification of a sensory 10.1016/j.stem.2012.07.006. PubMed: 22981823.

PLOS ONE | www.plosone.org 12 December 2013 | Volume 8 | Issue 12 | e83358 Coregulation by microRNAs and NR2F1

68. Chan CM, Fulton J, Montiel-Duarte C, Collins HM, Bharti N et al. (2013) proliferation, survival and migration of breast cancer cells. Breast A signature motif mediating selective interactions of BCL11A with the Cancer Res Treat 110: 69–83. doi:10.1007/s10549-007-9693-6. NR2E/F subfamily of orphan nuclear receptors. Nucleic Acids PubMed: 17674191. Research. doi:10.1093/nar/gkt761. 85. Rüsch A, Erway LC, Oliver D, Vennström B, Forrest D (1998) Thyroid 69. Zhang Z, Burch PE, Cooney AJ, Lanz RB, Pereira Fa et al. (2004) hormone receptor beta-dependent expression of a potassium Genomic analysis of the nuclear receptor family: new insights into conductance in inner hair cells at the onset of hearing. Proc Natl Acad structure, regulation, and evolution from the rat genome. Genome Res Sci U S A 95: 15758–15762. doi:10.1073/pnas.95.26.15758. PubMed: 14: 580–590. doi:10.1101/gr.2160004. PubMed: 15059999. 9861043. 70. Miyaki S, Sato T, Inoue A, Otsuki S, Ito Y et al. (2010) MicroRNA-140 86. Denver RJ, Williamson KE (2009) Identification of a thyroid hormone plays dual roles in both cartilage development and homeostasis. Genes response element in the mouse Kruppel-like factor 9 gene to explain its Dev 24: 1173–1185. doi:10.1101/gad.1915510. PubMed: 20466812. postnatal expression in the brain. Endocrinology 150: 3935–3943. doi: 71. Miyaki S, Nakasa T, Otsuki S, Grogan SP, Higashiyama R et al. (2009) 10.1210/en.2009-0050. PubMed: 19359381. MicroRNA-140 is expressed in differentiated human articular 87. Dettling J, Franz C, Zimmermann U, Lee SC, Bress A et al. (2013) chondrocytes and modulates interleukin-1 responses. Arthritis Rheum Autonomous functions of murine thyroid hormone receptor TRα and 60: 2723–2730. doi:10.1002/art.24745. PubMed: 19714579. TRβ in cochlear hair cells. Molecular and Cellular Endocrinology. doi: 72. Takata A, Otsuka M, Yoshikawa T, Kishikawa T, Hikiba Y et al. (2013) 10.1016/j.mce.2013.08.025. MicroRNA-140 acts as a liver tumor suppressor by controlling NF-κB 88. Cooney AJ, Tsai SY, O’Malley BW, Tsai MJ (1992) Chicken ovalbumin activity by directly targeting DNA methyltransferase 1 (Dnmt1) upstream promoter transcription factor (COUP-TF) dimers bind to expression. Hepatology 57: 162–170. doi:10.1002/hep.26011. PubMed: different GGTCA response elements, allowing COUP-TF to repress 22898998. hormonal induction of the vitamin D3, thyroid hormone, and retinoic 73. Liu Y, Zhang ZC, Qian SW, Zhang YY, Huang HY et al. (2013) acid receptors. Mol Cell Biol 12: 4153–4163. PubMed: 1324415. MicroRNA-140 promotes adipocyte lineage commitment of C3H10T1/2 89. Ying M, Sang Y, Li Y, Guerrero-Cazares H, Quinones-Hinojosa A et al. pluripotent stem cells via targeting osteopetrosis-associated (2011) Krüppel-like family of transcription factor 9, a differentiation- transmembrane protein 1. J Biol Chem 288: 8222–8230. doi:10.1074/ associated transcription factor, suppresses Notch1 signaling and jbc.M112.426163. PubMed: 23389033. inhibits glioblastoma-initiating stem cells. Stem Cells 29: 20–31. doi: 74. Yang J, Qin S, Yi C, Ma G, Zhu H et al. (2011) MiR-140 is co- 10.1002/stem.561. PubMed: 21280156. expressed with Wwp2-C transcript and activated by Sox9 to target Sp1 90. Abe S, Katagiri T, Saito-Hisaminato A, Usami S, Inoue Y et al. (2003) in maintaining the chondrocyte proliferation. FEBS Lett 585: 2992– Identification of CRYM as a candidate responsible for nonsyndromic 2997. doi:10.1016/j.febslet.2011.08.013. PubMed: 21872590. deafness, through cDNA microarray analysis of human cochlear and 75. Mak ACY, Szeto IYY, Fritzsch B, Cheah KSE (2009) Differential and vestibular tissues. Am J Hum Genet 72: 73–82. doi:10.1086/345398. overlapping expression pattern of SOX2 and SOX9 in inner ear PubMed: 12471561. development. Gene Expr Patterns 9: 444–453. doi:10.1016/j.gep. 91. Oshima A, Suzuki S, Takumi Y, Hashizume K, Abe S et al. (2006) 2009.04.003. PubMed: 19427409. CRYM mutations cause deafness through thyroid hormone binding 76. Spörl F, Korge S, Jürchott K, Wunderskirchner M, Schellenberg K et al. properties in the fibrocytes of the cochlea. J Med Genet 43: e25. doi: (2012) Krüppel-like factor 9 is a circadian transcription factor in human 10.1136/jmg.2005.034397. PubMed: 16740909. epidermis that controls proliferation of keratinocytes. Proc Natl Acad 92. Pingault V, Ente D, Dastot-Le Moal F, Goossens M, Marlin S et al. Sci U S A 109: 10903–10908. doi:10.1073/pnas.1118641109. PubMed: (2010) Review and update of mutations causing Waardenburg 22711835. syndrome. Hum Mutat 31: 391–406. doi:10.1002/humu.21211. 77. Pei H, Yao Y, Yang Y, Liao K, Wu J-R (2011) Krüppel-like factor KLF9 PubMed: 20127975. regulates PPARγ transactivation at the middle stage of adipogenesis. 93. Brown KK, Rehm HL (2012) Molecular diagnosis of hearing loss. Cell Death Differ 18: 315–327. doi:10.1038/cdd.2010.100. PubMed: Current protocols in human genetics / editorial board, Jonathan L 20725087. Haines . [et al.] Chapter 9 Unit9.16. doi: 78. Dai K, Khatun I, Hussain MM (2010) NR2F1 and IRE1beta suppress 10.1002/0471142905.hg0916s72. microsomal triglyceride transfer protein expression and lipoprotein 94. Davarnia B, Babanejad M, Fattahi Z, Nikzat N, Bazazzadegan N et al. assembly in undifferentiated intestinal epithelial cells. Arterioscler (2012) Spectrum of GJB2 (Cx26) gene mutations in Iranian Azeri Thromb Vasc Biol 30: 568–574. doi:10.1161/ATVBAHA.109.198135. patients with nonsyndromic autosomal recessive hearing loss. Int J PubMed: 20007910. Pediatr Otorhinolaryngol 76: 268–271. doi:10.1016/j.ijporl.2011.11.019. 79. Charitidi K, Meltser I, Canlon B (2012) Estradiol treatment and PubMed: 22172221. hormonal fluctuations during the estrous cycle modulate the expression 95. Krishnan R, Cleary EG (1990) Elastin gene expression in elastotic of estrogen receptors in the auditory system and the prepulse inhibition human breast cancers and epithelial cell lines. Cancer Res 50: 2164– of acoustic startle response. Endocrinology 153: 4412–4421. doi: 2171. PubMed: 2156612. 10.1210/en.2012-1416. PubMed: 22778224. 96. Lapis K, Tímár J (2002) Role of elastin-matrix interactions in tumor 80. Simonoska R, Stenberg AE, Duan M, Yakimchuk K, Fridberger A et al. progression. Semin Cancer Biol 12: 209–217. doi:10.1016/ (2009) Inner ear pathology and loss of hearing in estrogen receptor- S1044-579X(02)00024-X. PubMed: 12083851. beta deficient mice. J Endocrinol 201: 397–406. doi:10.1677/ 97. Fu L, Balasubramanian M, Shan J, Dudenhausen EE, Kilberg MS JOE-09-0060. PubMed: 19293293. (2011) Auto-activation of c-JUN gene by amino acid deprivation of 81. Meltser I, Tahera Y, Simpson E, Hultcrantz M, Charitidi K et al. (2008) hepatocellular carcinoma cells reveals a novel c-JUN-mediated Estrogen receptor beta protects against acoustic trauma in mice. J Clin signaling pathway. J Biol Chem 286: 36724–36738. doi:10.1074/ Invest 118: 1563–1570. doi:10.1172/JCI32796. PubMed: 18317592. jbc.M111.277673. PubMed: 21862593. 82. Simmons CD, Pabona JMP, Zeng Z, Velarde MC, Gaddy D et al. 98. Vogt PK (2001) Jun, the oncoprotein. Oncogene 20: 2365–2377. doi: (2010) Response of adult mouse uterus to early disruption of estrogen 10.1038/sj.onc.1204443. PubMed: 11402333. receptor-alpha signaling is influenced by Krüppel-like factor 9. J 99. Inoue K, Shiga T, Ito Y (2008) Runx transcription factors in neuronal Endocrinol 205: 147–157. doi:10.1677/JOE-09-0474. PubMed: development. Neural Dev 3: 20. doi:10.1186/1749-8104-3-20. PubMed: 20164373. 18727821. 83. Zhang Y, Eades G, Yao Y, Li Q, Zhou Q (2012) Estrogen receptor α 100. Eberhart JK, He X, Swartz ME, Yan Y-L, Song H et al. (2008) signaling regulates breast tumor-initiating cells by down-regulating MicroRNA Mirn140 modulates Pdgf signaling during palatogenesis. Nat miR-140 which targets the transcription factor SOX2. J Biol Chem 287: Genet 40: 290–298. doi:10.1038/ng.82. PubMed: 18264099. 41514–41522. doi:10.1074/jbc.M112.404871. PubMed: 23060440. 84. Le Dily F, Métivier R, Guéguen M-M, Le Péron C, Flouriot G et al. (2008) COUP-TFI modulates estrogen signaling and influences

PLOS ONE | www.plosone.org 13 December 2013 | Volume 8 | Issue 12 | e83358